A computer-implemented voice-based analysis system

Abstract

The present invention discloses a computer-implemented voice-based analysis system (100) comprising: a voice recording unit (110) configured to receive voice-based audio inputs associated with a user; a data processing module (120) configured to process the voice-based audio inputs into processed data; a machine learning algorithm (130) executed by a processor, being configured to analyse the processed data and generate personalised recommendations for specific purposes; a database (140) storing a library of compounds and their corresponding therapeutic properties; a compound selection module (150) configured to access the database and select one or more compounds from the database based on the personalised recommendations generated by the machine learning algorithm; and a compounding unit (160) configured to combine the selected compounds in predetermined quantities to form a purpose compounding supplement; wherein the voice-based analysis system (100) is configured to receive and analyse audio data associated with a user to determine mental stress level, sleep quality, and overall mental health of the user.

Description

[0001] A COMPUTER-IMPLEMENTED VOICE-BASED ANALYSIS SYSTEM

[0002] FIELD OF INVENTION

[0003] The invention relates generally to the fields of computer-implemented analysis system. More particularly, the invention relates to a computer-implemented voice-based analysis system for determining mental stress level, sleep quality, and overall mental health of a user.

[0004] BACKGROUND OF THE INVENTION

[0005] Mental health plays a vital role in our overall well-being, influencing our emotions, thoughts, and behaviours. If affects how we cope with stress, interact with others, and make choices in our daily lives. However, despite its significance, mental health concerns often go unnoticed or unaddressed. Studies indicate that approximately one in four individuals will experience a mental health issue during their lifetime. These conditions can range from common disorders such as anxiety and depression to more severe conditions like bipolar disorder or schizophrenia. The impact of mental health problems can be profound, affecting various aspects of a person’s life, including their relationships, work performances, and overall quality of life.

[0006] In the field of mental health assessment and recommendations, traditional approaches have often relied on subjective evaluations or questionnaire-based methods. These methods are often time-consuming and can be influenced by biases, resulting in less accurate assessments and recommendations. As technology has advanced, there has been a growing recognition of the need for automated and efficient systems that can provide objective and personalized evaluations of mental health.

[0007] Voice analysis has emerged as a promising approach in this regard. It is well-known that an individual’s vocal patterns, including pitch, tone, rhythm, and other acoustic characteristics, can be influenced by their mental health status. Research has shown that changes in vocal parameters can be indicative of mental health conditions such as depression, anxiety, and stress. By leveraging voice analysis, a system can capture and analyse objective data from an individual’s vocal patterns, providing valuable insights into their mental health. Unlike traditional assessment methods, voice analysis offers a non-invasive and unobstructive approach, as it does not require physical sensors or invasive procedures. Instead, it utilizes the existing technology of voice recording units to capture voice-based audio inputs associated with the user.

[0008] Generally, the use of audio and frequency for detecting mental health status has been previously explored in research studies. For instance, Schuller et al. (2010) and Cummins et al. (2013) conducted studies demonstrating that voice pattern analysis can be effective in detecting depression and anxiety in individuals. These studies highlighted the potential of utilizing voice-based analysis as a tool for objective mental health assessment.

[0009] Furthermore, research conducted by Kim et al. (2019) showed that acoustic analysis can be utilized to detect sleep disorders. This study highlighted the feasibility of using voice-based audio inputs to assess sleep quality, providing additional evidence of the potential of voice analysis in mental health evaluation. Another relevant study by Graesser et al. (2011) explored the use of acoustic analysis to detect concentration levels. The research demonstrated the relationship between voice characteristics and an individual’s cognitive state, providing further support for the applicability of voice analysis in mental health assessment.

[0010] However, it is important to note that previous studies primarily relied on manual analysis method and did not incorporate machine learning algorithms or automated formulation of compounding supplements based on the collected data. Accordingly, it would be desirable to provide a method which is able to overcome the aforesaid drawbacks. This invention provides such a method.

[0011] SUMMARY OF INVENTION

[0012] The main objective of the invention is to provide users with a computer-implemented voice-based analysis system. Preferably, embodiments of the present disclosure may include a voice recording unit configured to receive voice-based audio inputs associated with a user, a data processing module configured to process the voice-based audio inputs into processed data, a machine learning algorithm executed by a processor, being configured to analyse the processed data and generate personalised recommendations for specific purposes, a database storing a library of compounds and their corresponding therapeutic properties, a compound selection module configured to access the database and select one or more compounds from the database based on the personalised recommendations generated by the machine learning algorithm, a compounding unit configured to combine the selected compounds in predetermined quantities to form a purpose compounding supplement.

[0013] Preferably, the voice-based analysis system may be configured to receive and analyse audio data associated with a user to determine mental stress level, sleep quality, and overall mental health of the user.

[0014] Embodiments of the present disclosure may also include a feature extraction module configured to extract one or more non-lexical features from the voice-based audio inputs. In some embodiments, the non-lexical features include articulation space, space, envelope peaks, vocal effort and the combination thereof.

[0015] Embodiments of the present disclosure may also include a predictive model trained using a plurality of voice-based audio inputs and one or more non-lexical features, to determine the mental stress level, sleep quality, and overall mental health of the user. Embodiments of the present disclosure may also include a probability calculation module configured to calculate the probability of the individual experiencing mental stress, sleep disturbances or poor mental health based on the extracted non-lexical features and the outcome of predictive model.

[0016] Embodiments of the present disclosure may also include an output module configured to provide the calculated probabilities to the data processing module.

[0017] Embodiments of the present disclosure may also include a comparison module configured to compare the processed data with reference data or established benchmarks related to mental stress level, sleep quality, and overall mental health.

[0018] Embodiments of the present disclosure may also include a dosage optimization module configured to optimise the quantities of selected compounds based the personalised recommendations and individual characteristics.

[0019] Embodiments of the present disclosure may also include a compound substitution module configured to suggest alternative compounds with similar therapeutic properties in case certain compounds may be unavailable or contraindicated for the user.

[0020] Embodiments of the present disclosure may also include a formulation generation module configured to generate a formulation recipe specifying the quantities and combination of selected compounds for the purpose compounding supplement.

[0021] References throughout this specification to features, advantages, or similar language do not imply that all of the features and advantages may be realized in any single embodiment. Rather, language referring to the features and advantages is understood to mean that a specific feature, advantage, or characteristic is included in at least one embodiment. Thus, discussion of the features and advantages, and similar language, throughout this specification may, but do not necessarily, refer to the same embodiment.

[0022] Furthermore, the described features, advantages, and characteristics of the embodiments may be combined in any suitable manner. One skilled in the relevant art will recognize that the embodiments may be practiced without one or more of the specific features or advantages of a particular embodiment. In other instances, additional features and advantages may be recognized in certain embodiments that may not be present in all embodiments.

[0023] One skilled in the art will readily appreciate that the invention is well adapted to carry out the objects and obtain the ends and advantages mentioned, as well as those inherent therein. The embodiments described herein are not intended as limitations on the scope of the invention.

[0024] BRIEF DESCRIPTION OF THE DRAWINGS

[0025] For the purpose of facilitating an understanding of the invention, there is illustrated in the accompanying drawing the preferred embodiments from an inspection of which when considered in connection with the following description, the invention, its construction and operation and many of its advantages would be readily understood and appreciated.

[0026] Figure 1 is a block diagram that describes a computer-implemented voice-based analysis system.

[0027] Figure 2 IS a block diagram that further describes the computer-implemented voice-based analysis system, according to some embodiments of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0028] The invention will now be described in greater detail, by way of example, with reference to the drawings.

[0029] The following detailed description of embodiments of the present technology references the accompanying drawings that illustrate specific embodiments in which the technology can be practiced. The embodiments are intended to describe aspects of the technology in sufficient detail to enable those skilled in the art to practice the technology. Other embodiments can be utilized and changes can be made without departing from the scope of the present technology. The following detailed description is, therefore, not to be taken in a limiting sense. The scope of the present technology is defined only by the appended claims, along with the full scope of equivalents to which such claims are entitled.

[0030] For example, the steps recited in any of the method or process descriptions may be executed in any order and are not limited to the order presented. Moreover, any of the functions or steps may be outsourced to or performed by one or more third parties. Furthermore, any reference to singular includes plural embodiments, and any reference to more than one component may include a singular embodiment.

[0031] Reference throughout this specification to “one embodiment”, “an embodiment,” or similar language means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present subject matter. Thus, appearance of the phrases “in one embodiment,” “in an embodiment,” and similar language throughout this specification may, but do not necessarily, all refer to the same embodiment.

[0032] Unless otherwise explained, all technical and scientific terms used herein have the sam meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The singular terms “a”, “an”, and “the” include plural referents unless context clearly indicates otherwise. Similarly, the word “or” is intended to include “and” unless the context clearly indicates otherwise. Hence “comprising A or B” means including A, or B, or A and B. It is further to be understood that all base sizes or amino acid sizes, and all molecular weight or molecular mass values, given for nucleic acids or polypeptides are approximate, and are provided for description. The terms “about”, “substantially” and “approximately” mean, in general, the stated value plus or minus 5%.

[0033] Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present disclosure, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirely. In case of conflict, the present specification, including explanations of terms, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

[0034] FIG. l is a block diagram that describes a computer-implemented voice-based analysis system 100. The voice-based analysis system 100 may be configured to receive and analyse audio data associated with a user to determine mental stress level, sleep quality, and overall mental health of the user.

[0035] In some embodiments, the computer-implemented voice-based analysis system 100 may include a voice recording unit 110 configured to receive voice-based audio inputs associated with a user. Preferably, the voice recording unit 110 refers to a component that is specifically designed to capture and receive voice-based audio inputs associated with the user. The voice recording unit 110 serves as the interface between the user and the system, allowing the system to gather the necessary audio data for further processing and analysis. When a user interacts with the system, the voice recording unit 110 activates and starts recording the user’s voice. This can be done various means, including but not limited to, a microphone, a dedicated recording devise, or an application running on a device such as smartphone or computer. The voice-based audio inputs received by the voice recording unit 110 may include various types of user speech, such as commands, queries, responses, or any other form of vocal communication. The voice-based audio inputs may also contain spoken words, expressions, sound, or non-verbal vocal cues. Preferably, the voice recording unit 110 ensures that the voice-based audio inputs are captured accurately and in a suitable format for subsequent analysis.

[0036] In some embodiments, the computer-implemented voice-based analysis system 100 may include a data processing module 120 configured to process the voice-based audio inputs into processed data. Preferably, the data processing module 120 refers to a component designed to take raw voice-based audio inputs and various processing techniques to transform the raw voice-based audio inputs into processed data. Preferably, the processing techniques include the step of initial cleaning the preparation of the raw voice-based audio inputs, which may include tasks such as noise reduction, removal of background sounds, normalisation of audio levels, and filtering to eliminate artifacts or disturbances. Next, the data processing module 120 may extracts relevant features or characteristics form the audio inputs to capture important aspects of the audio inputs for further analysis. One the features are extracted, the data processing module 120 may perform additional transformations of the extracted features, which may involve normalizing the feature value, applying statistical techniques, or converting the data into a suitable representation for subsequent analysis.

[0037] In some embodiments, the computer-implemented voice-based analysis system 100 may include a machine learning algorithm 130 executed by a processor, being configured to analyse the processed data and generate personalise recommendations for specific purposes. Preferably, the machine learning algorithm 130 includes a computational model that is designed to analyse the processed data obtained from the voice-based audio inputs. The computation model leverages machine learning techniques to automatically learn patterns, correlations, and relationships within the data and generate personalized recommendation for specific purposes. Preferably, the processed data with various features serves as input to the machine learning algorithm 130, whereby the machine learning algorithm 130 processes the processed data features representation and analyses it to capture the relevant information for generating recommendations. Further, the machine learning algorithm 130 employs statistical techniques and algorithms to learn patterns, correlations, and dependencies within the processed data, to identifies relationships between the features and the target recommendations based on a labelled dataset provided during training phrase. Once the machine learning algorithm 130 has learned from the training data, it can generate personalised recommendations for specific purposes. By iteratively adjusting its internal parameters, the machine learning algorithm 130 optimises its ability to make accurate predictions or recommendations. Furthermore, the machine learning algorithm’s 130 recommendations can be further refined and improved over time by continuously feeding it with new data and updating its model. This enables the machine learning algorithm 130 to adapt to changes in user preferences and provide increasingly accurate and personalised recommendations.

[0038] In some embodiments, the computer-implemented voice-based analysis system 100 may include a database 140 storing a library of compounds and their corresponding therapeutic properties. Preferably, the database 140 serves as a repository of information about various chemical compounds and their associated therapeutic effects. Preferably, the database 140 stores detailed information about different chemical compounds, including their chemical structures, molecular properties, physiochemical characteristics, and other relevant identifiers. This information allows for precise identification and classification of compounds. Preferably, the database 140 associates each compound with its specific therapeutic properties, which can include indications, mechanisms of action, target molecules, disease states, clinical trial data, and any other relevant information related to the compound’s potential therapeutic benefits. Preferably, the database 140 may organise compounds into various categories or classifications based on their chemical structure, therapeutic class, or other relevant criteria. This classification allows for easier navigation and search within the database 140, enabling the machine learning algorithm 130 to find compounds with specific therapeutic properties more efficiently.

[0039] Alternatively, the database 140 may also provides search and retrieval functionalities that allow the machine learning algorithm 130 to query and retrieve compounds based on specific criteria, such as therapeutic properties, chemical structure, or compound identifiers. The machine learning algorithm 130 can perform keyword searches, apply filters or use advanced search techniques to find compounds of interest.

[0040] Alternatively, the database 140 may also integrate data from multiple sources, such as scientific literature, clinical trials, or proprietary research, to provide a comprehensive and up-to-date repository of compound information. Regular updates ensure that the database 140 reflects the latest findings and developments in the field of therapeutic compounds.

[0041] Alternatively, the database 140 may also offer analytical tools and visualization capabilities to facilitate data exploration and analysis, the machine learning algorithm 130 can perform data mining, statistical analysis, or generate visual representations to identify patterns, trends, or relationships among compounds and their therapeutic properties.

[0042] In some embodiments, the computer-implemented voice-based analysis system 100 may include a compound selection module 150 configured to access the database 140 and select one or more compounds from the database 140 based on the personalised recommendations generated by the machine learning algorithm 130. Preferably, the compound selection module 150 establishes a connection to the database 140 and utilises query mechanisms to access the compound data and associated therapeutic properties. Further, the compound selection module 150 utilises the personalised recommendations generated by the machine learning algorithm 130 as input for compound selection. These recommendations are based on the analysis of processed data from voice-based audio inputs, taking into account factors such as the user’s preferences, medical history, genetic profile, or other relevant information. Preferably, the compound selection module 150 utilises the personalised recommendations to define specific criteria for selection compounds from the database 140. These criteria can include, but not limited to, desired therapeutic properties, target molecules, indications, or any other relevant parameters identified by the machine learning algorithm 130. By using the defined matching criteria, the compound selection module 150 performs a search and selection process within the database to identify one or more compounds that meet the personalised recommendations. Preferably, the compound selection module 150 may utilise querying mechanisms, filtering techniques, or ranking algorithms to identify the most relevant compounds base on their therapeutic properties and associated data. Once the compound selection process is complete, the compound selection module 150 generates a list of recommended compounds that match the user’s personalised requirements. Preferably, these recommendations can include information about the selected compounds, such as their names, therapeutic properties, chemical structures, or any other relevant details that aid the user in understanding their potential benefits. Alternatively, the compound selection module 150 may establish a feedback loop with the machine learning algorithm 130 to continuously improve the selection process. Preferably, user feedback on the recommended compounds, such as effectiveness or side effects experienced, can be collected and utilised to refine the machine learning algorithm’s recommendations and enhance future compound selection. Preferably, the compound selection module 150 further includes a comparison module 151, configured to compare the processed data with reference data or established benchmarks related to mental stress level, sleep quality, and overall mental health. This comparison module 151 allows for a comprehensive assessment of the individual's condition by comparing their data with known standards or benchmarks in the field. Preferably, The comparison module 151 has access to reference data or established benchmarks that serve as standards or comparative measures for mental stress level, sleep quality, and overall mental health. These reference data or benchmarks can be obtained from clinical studies, research databases, expert guidelines, or established norms in the field. The comparison module 151 compares the processed data of the individual with the reference data or established benchmarks. It evaluates how the individual's data aligns or deviates from these established standards. This comparison can involve statistical analysis, pattern matching, or other comparative techniques. Overall, the comparison module 151 serves as a decision support tool, providing valuable information to guide the compound selection process. The comparison results can influence the selection of compounds based on the individual's specific needs and conditions. For example, if the comparison reveals a higher-than-normal mental stress level, the module may prioritize compounds with known properties for stress reduction or relaxation. Lastly, the comparison module 151 contributes to the generation of personalized recommendations by taking into account the comparison results. It helps in identifying the most relevant compounds or interventions based on the individual's specific condition, ensuring a targeted approach to addressing mental stress, sleep quality, and overall mental health.

[0043] In some embodiments, the computer-implemented voice-based analysis system 100 may include a compounding unit 160 configured to combine the selected compounds in predetermined quantities to form a purpose compounding supplement. Preferably, the compounding unit 160 includes a formulation generation module 161 that is specifically designed to generate a formulation recipe. Preferably, the recipe may specifies the quantities and combination of the selected compounds to be used in the purpose compounding supplement. Preferably, the formulation generation module 161 receives input from the compound selection module 150, which includes the list of selected compounds and their respective quantities or concentrations. Preferably, the compound selection module 150 utilises algorithms and rules to generate a formulation recipe that indicates the precise quantities and combination of the selected compounds. These algorithms may consider factors such as therapeutic properties of the compounds, their compatibility, stability, solubility, and any specific requirements for compounding. The formulation generation module 161 may also performs proportional calculations to determine the appropriate quantities of each selected compound based on the desired formulation, taking into account the desired dosage, concentration, or strength of the compounds to achieve the desired therapeutic effect. Preferably, the formulation generation module 161 may also conducts compatibility checks to ensure that the selected compounds can be safely combined and that there are no interactions or adverse effects when mixed, by considering factors such as chemical compatibility, stability, pH requirements, and any potential incompatibilities between the compounds. Preferably, the formulation generation module 161 may generates a detailed formulation recipe that specifies the quantities and combination of the selected compounds. The generated detailed formulation recipe may include the exact measurements, proportions, and any specific instructions or steps required for compounding, which serves as a guide for the compounding unit 160 to accurately combine the compounds.

[0044] Alternatively, generated detailed formulation recipe may undergo validation and verification processes to ensure its accuracy and suitability for the intended purpose by a dosage optimisation module 162. Preferably, the dosage optimisation module 162 takes into account various factors such as personalised recommendations generated by the machine learning algorithm 130 and individual characteristics to determine the most appropriate dosage for the selected compounds, ensuring an optimized and tailored approach to supplementation. Further, the dosage optimization module 162 may also utilizes algorithms or rules to determine the optimal quantities of the selected compounds. These algorithms take into account the personalized recommendations, individual characteristics, and known dosage guidelines or standards for the compounds. The dosage optimisation module 162 may employ an iterative process to refine the dosage optimization. It may consider feedback from the individual, healthcare professionals, or further analysis of the individual's response to the supplementation. Lastly, the dosage optimisation module 162 provides documentation and instructions specifying the optimized quantities of the selected compounds for the compounding unit. This information ensures accurate and consistent formulation of the purpose compounding supplement.

[0045] In some alternative embodiment, the compound unit 160 includes a compound substitution module 163, which is designed to suggest alternative compounds with similar therapeutic properties in the event that certain compounds are unavailable or contraindicated for the user. Preferably, the compound substitution module 163 evaluates the availability of the selected compounds in real-time and checks whether the compounds are currently in stock or accessible within the compounding unit's inventory. Additionally, the compound substitution module 163 may also assesses any contraindications or specific restrictions that may prevent the use of certain compounds for the user's health condition, allergies, or other factors. Further, the compound substitution module 163 may leverages an alternative compound database that stores information about compounds with similar therapeutic properties to those initially selected. Preferably, The compound substitution module 163 uses matching criteria and substitution rules to determine suitable alternative compounds. These criteria consider factors such as the therapeutic properties, pharmacological profiles, chemical structures, and known interactions of the original compounds. The substitution rules ensure that the suggested alternatives possess similar therapeutic potential to the original compounds. Based on the availability and contraindication assessment, as well as the matching criteria and substitution rules, the compound substitution module 163 suggests alternative compounds to replace any unavailable or contraindicated compounds. These suggestions provide options that are expected to deliver similar therapeutic effects as the original compounds, ensuring the continuity of treatment or desired outcomes. Alternatively, the compound substitution module 162 may provide information about the suggested alternative compounds to the user or healthcare professional involved in the compounding process.

[0046] FIG. 2 is a block diagram that further describes the computer-implemented voice-based analysis system 100 from FIG. 1, according to some embodiments of the present disclosure. In some embodiments, the computer-implemented voice-based analysis system 100 as claimed, further includes a predictive model 131 incorporates a feature extraction module 220 configured to extract one or more non-lexical features 230 from the voice-based audio inputs.

[0047] As used herein, the term “feature extraction module” 220 refers to a component designed to process and analyse voice-based audio input in order to extract specific characteristics or attributes that are not related to the actual words or content of the speech. These extracted features provide additional information about the audio inputs, allowing for further analysis and understanding of the audio data.

[0048] In a preferred embodiment, non-lexical features 230 refer to the aspects of speech that are not directly related to the words being spoken. Preferably, non-lexical features 230 focus on the manner in which speech is produced, rather than the specific linguistic content. By analysing these non-lexical features, valuable insights can be gained about the user, the emotional state, or the overall acoustic properties of the audio. In some embodiments as according to the present disclosure, the non-lexical features 230 may include pitch and intonation 232, tempo and rhythm 234, voice quality 236, spectral features 238, energy and loudness 240, and the combination thereof.

[0049] Preferably, pitch and intonation 232 relate to the variation in the fundamental frequency of the voice-based audio inputs, providing information about the user’s emphasis, mood, or emotional state.

[0050] Preferably, tempo and rhythm 234 refers to the speed and pattern of speech, which can indicate factors like the user’s level of confidence or the presence of hesitation or stress.

[0051] Preferably, voice quality 236 focuses on the overall characteristics of the voice-based audio inputs, including measures of clarity, hoarseness, breathiness, or roughness, which can provide insights into the user’s physical or emotional state.

[0052] Preferably, spectral features 238 capture information about the frequency distribution of the voice-based audio inputs, such as formants or resonance peaks, which can indicate the user’s gender, age or even the presence of certain medical conditions.

[0053] Preferably, energy and loudness 240 relate to the overall volume and intensity of the voice-based audio inputs, which can reflect the user’s level of excitement, engagement, or fatigue.

[0054] In some embodiment, the machine learning algorithm 130 may also include a probability calculation module 132 that is configured to calculate the probability of an individual experiencing mental stress, sleep disturbances, or poor mental health. Preferably, the probability calculation module 132 utilizes the extracted non-lexical features from the voice-based audio inputs and the outcomes generated by the predictive model 131 to estimate the likelihood of these mental health conditions. These outcomes can be represented as probabilities or confidence scores associated with each mental health condition. Using the extracted non-lexical features 230 and the outcomes from the predictive model, the probability calculation module 132 applies statistical techniques or algorithms to estimate the probability of the individual experiencing mental stress, sleep disturbances, or poor mental health. These calculations consider the relationships between the non-lexical features 230 and the predicted mental health indicators learned by the predictive model 131 during the training process. The probability calculation module 132 may interpret the calculated probabilities based on predefined thresholds or criteria. Thresholds define the boundary values above which a specific mental health condition is considered likely. By calculating the probabilities of mental health conditions, the probability calculation module 132 enables a risk assessment for the individual. This assessment provides insights into the likelihood and severity of mental stress, sleep disturbances, or poor mental health. Lastly, feedback is generated by a output module 133 based on these probabilities for used in guiding interventions, recommendations, or further assessments to support the individual's mental well-being.

[0055] In a first exemplary embodiment, the present disclosed system can be applied to detect stress levels by analysing the acoustic features of a recorded audio signal. Preferably, a recorded audio signal is processed to extract relevant acoustic features such as pitch, amplitude, and duration. These features capture variations in the individual's voice that may reflect their stress level. The extraction process involves analysing the audio signal and quantifying these acoustic characteristics. The extracted acoustic features are input into a machine learning algorithm 130, which has been trained on a dataset of labelled audio samples with corresponding stress levels. The machine learning algorithm 130 learns patterns and relationships between the acoustic features and the stress levels to build a predictive model 131. Further, The machine learning algorithm 130 utilizes the extracted acoustic features and the trained predictive model 131 to predict the stress level of the individual. By analysing the acoustic features in relation to the patterns learned during training, the machine learning algorithm 130 can estimate the stress level with a certain degree of accuracy. The predicted stress level can be displayed in a graphical form, providing a visual representation of the individual's stress level over time. This graphical representation may include a timeline or trend chart illustrating the fluctuations or changes in the stress level. It allows the individual or a healthcare professional to monitor and track stress levels and identify patterns or triggers. Based on the predicted stress level, the system can automatically formulate a compounding supplement tailored to reducing stress levels. This involves utilizing the predicted stress level as input for a compounding unit 160, which combines selected compounds in predetermined quantities to create a supplement specifically targeted at stress reduction. The formulation process can be guided by predefined recipes or algorithms that determine the appropriate compounds and their proportions based on the predicted stress level.

[0056] In a second exemplary embodiment, the present disclosed system can be applied to detect sleep quality by analysing the acoustic features of a recorded audio signal during sleep. Preferably, the recorded audio signal captured during sleep is processed to extract relevant acoustic features, such as changes in frequency and intensity. These features capture variations in the audio signal that are indicative of the individual's sleep patterns and quality. The extracted acoustic features during sleep are input into a machine learning algorithm 130, which has been trained on a dataset of labelled audio samples with corresponding sleep quality indicators. The machine learning algorithm 130 learns patterns and relationships between the acoustic features and the sleep quality indicators to build a predictive model 131. The machine learning algorithm 130 utilizes the extracted acoustic features and the trained predictive model 131 to predict the sleep quality of the individual. By analysing the acoustic features in relation to the patterns learned during training, the machine learning algorithm 130 can estimate the sleep quality with a certain degree of accuracy. Preferably, the predicted sleep quality can be displayed in a graphical form, providing a visual representation of the individual's sleep quality over time. This graphical representation may include a timeline or trend chart illustrating the variations or changes in sleep quality. It allows the individual or a healthcare professional to monitor and track sleep quality and identify patterns or factors affecting sleep. Based on the predicted sleep quality, the system can automatically formulate a compounding supplement tailored to improving sleep quality. This involves utilizing the predicted sleep quality as input for a compounding unit 160, which combines selected compounds in predetermined quantities to create a supplement specifically targeted at improving sleep. The formulation process can be guided by predefined recipes or algorithms that determine the appropriate compounds and their proportions based on the predicted sleep quality.

[0057] It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession way, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. Other steps and methods may be conceived that are equivalent in function, logic, or effect to one or more blocks, or portions thereof, of the illustrated figures.

[0058] The present disclosure includes as contained in the appended claims, as well as that of the foregoing description. Although this invention has been described in its preferred form with a degree of particularity, it is understood that the present disclosure of the preferred form has been made only by way of example and that numerous changes in the details of construction and the combination and arrangements of parts may be resorted to without departing from the scope of the invention.

Claims

CLAIMS1. A computer-implemented voice-based analysis system (100) comprising: a voice recording unit (110) configured to receive voice-based audio inputs associated with a user; a data processing module (120) configured to process the voice-based audio inputs into processed data; a machine learning algorithm (130) executed by a processor, being configured to analyse the processed data and generate personalised recommendations for specific purposes; a database (140) storing a library of compounds and their corresponding therapeutic properties; a compound selection module (150) configured to access the database and select one or more compounds from the database based on the personalised recommendations generated by the machine learning algorithm; and a compounding unit (160) configured to combine the selected compounds in predetermined quantities to form a purpose compounding supplement; wherein the voice-based analysis system (100) is configured to receive and analyse audio data associated with a user to determine mental stress level, sleep quality, and overall mental health of the user.

2. The computer-implemented voice-based analysis system (100) as claimed in claim 1, further comprising a feature extraction module (220) configured to extract one or more non-lexical features (230) from the voice-based audio inputs.

3. The computer-implemented voice-based analysis system as claimed in claim 2, wherein the non-lexical features (230) include pitch and intonation (232), tempo and rhythm (234), voice quality (236), spectral features (238), energy and loudness (240), and the combination thereof.

4. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the machine learning algorithm (130) further comprising a predictive model (131) trained using a plurality of voice-based audio inputs and one or more non-lexical features, to determine the mental stress level, sleep quality, and overall mental health of the user.

5. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the machine learning algorithm (130) further comprising a probability calculation module (132) configured to calculate the probability of the individual experiencing mental stress, sleep disturbances or poor mental health based on the extracted non-lexical features and the outcome of predictive model (131).

6. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the machine learning algorithm (130) further comprising an output module (133) configured to provide the calculated probabilities feedback.

7. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the compound selection module (150) further comprising a comparison module (151) configured to compare the processed data with reference data or established benchmarks related to mental stress level, sleep quality, and overall mental health.

8. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the compounding unit (160) further comprising a formulation generation module (161) configured to generate a formulation recipe specifying the quantities and combination of selected compounds for the purpose compounding supplement.

9. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the compounding unit (160) further comprising a dosage optimization module (162) configured to optimise the quantities of selected compounds based the personalised recommendations and individual characteristics.

10. The computer-implemented voice-based analysis system (100) as claimed in any one of the preceding claims, wherein the compounding unit (160) further comprising a compound substitution module (163) configured to suggest alternative compounds with similar therapeutic properties in case certain compounds are unavailable or contraindicated for the user.

Images