Novel approach to early predict and diagnosis mental health issues in workplacesusing Artificial Intelligence (AI)

Abstract

A method for predicting early onset of mental health issues in work environments using artificial intelligence (AI) comprising: a) collecting data from two sources: (1) derived from employee survey an

Claims

1. A method for predicting and diagnosing early onset of mental health issues in work environments using artificial intelligence (Al), comprising:a). collecting data from two sources: (1) derived from employee survey. The data derived from this process represents a reference information at the time of survey (or static data), and (2) metadata collected from employee’s day-to-day activities i.e., digital footprint. The metadata is collected on a continual basis (dynamic data) and used to provide a dynamic insight into the employee’s day-to-day activity.b). analysing the data from both sources for each employee to identify potential clinical condition such as Stress, Mood, Anxiety, Depression etc.c). identifying a root cause for any early onset of mental health condition identified based at least in part on the collected user data. This is so that preventative action can be undertaken by the employer; andd). providing the employee with an ability to engage in confidential discussions with a virtual psychiatrist (or a counsellor).

2. The method of claim 1, wherein collecting static data involves using a smart survey tool which asks the employee a series of specific multiple-choice questions and associated textual comments.

3. The method of claim 1, wherein collecting dynamic data involves extracting metadata from employee’s digital footprint. The use of work-based software packages and tools by employees as part of everyday work naturally creates a digital footprint such as time of use, types of activities etc. Whilst some of this information is confidential and could be legally privileged, it is important to ensure that only generalised or anonymised data (or metadata) is extracted.

4. The method of claim 1, wherein both static and dynamic data are analysed using Predictive Al techniques. The Predictive Al models are implemented using software code and for each type of clinical condition (such as stress, anxiety, depression etc) there is a dedicated Predictive Al model.

5. The method of claim 4, wherein Predictive Al model comprises of supervised machine learning (ML) / deep learning (DL) classifiers. The classifier’s main goal is to predict an accurate scale answer. This answer is adjusted in line with clinical guidelines to deduces accurate analytics regarding employee’s work-related mental health issues.

6. The method of claim 5, wherein several classifiers are used, with each classifier designed to analyse a specific clinical condition such Stress, Mood, Anxiety, Depression etc.

7. The method of claim 5, wherein information from the classifier is analysed and presented to the human resources (HR) team via the communication dashboard showing overall state of company’s mental-health and wellbeing status. This communication dashboard also highlights areas (teams or individuals) that could be under pressure (work-related or personal). Whilst the exact reasons for this anomaly is not available to HR, it nevertheless enables the HR team to take preventative measures to ensure that the employee does not experience “burn-out” or mental breakdown.

8. The method of claim 1, wherein comprises of a chatbot to enables the employee to engage in confidential discussions with a virtual psychiatrist (or a counsellor). This can help individuals understand symptoms and treatment options for various conditions. By recognising early signs of mental health issues, employees can seek care before conditions worsen.

9. The method of claim 8, wherein the chatbot comprises of deep learning Generative Al (GenAI) models. The GenAI model takes input from workplace metadata, text data, and survey scores related to mental health indicators (predictive model outputs). GenAI uses this data to analyse and identify patterns, it is also used to understand symptoms and explore treatment options for various conditions starting with anxiety, mood disorders, depression, and work-related stress. It automates manual tasks to tailor individuals in the best way that suits their personality and signs and symptoms of work-related stress or mental health issues and potentially reduce overall work hours.

10. A system for predicting early onset of mental health conditions in work environment, comprising:a). a processor.b). a memory holding instructions that, when executed by the processor, cause the processor to:i) continually collect and extract metadata from user’s digital footprint.ii) save the metadata in a database comprising user activities.iii) algorithmically analyse the saved data against baseline pattern obtained through employee survey.iv) identify any outlier patterns that that could indicate potential onset of mental health condition.v) provide the ability for a user to engage in confidential discussions with a virtual psychiatrist or counsellor.

11. The system of claim 10, wherein said the electronic device consists of at least one or more of the following: personal computer, mobile phone, and a laptop.

12. The system of claim 10, further comprising of a remote or cloud server in electronic communication with the said system, and said server comprises at least one of: a database of user data associated with a plurality of known users.

13. The system of claim 10, wherein said server algorithmically analyses the user data patterns to identify any outliers that could point to early onset of one or more clinical mental health conditions.

14. The system of claim 10, wherein the virtual psychiatrist or counsellor is a chatbot utilising at least one of an artificial intelligence (Al) based dialogue conversational engine, or Al based mental health and wellbeing response engine.

15. The system of claim 10, wherein said processor executes said instruction automatically without requiring continuous user activity.

16. The system of claim 10, comprising of the following components: an algorithmic processing unit consisting of artificial intelligence (Al) models for identifying user mental health conditions; a service fapade for soliciting user surveys; a data collecting system for extracting baseline (static) data and metadata (dynamic); a reporting service component for issuing mental health alerts to employer and associated clinical support staff; and a service fapade for the user to engage in confidential discussions with a virtual psychiatrist or counsellor.

Images