PARIS--(BUSINESS WIRE)--MyndBlue, a digital MedTech company specializing in major depressive disorder (MDD), post-traumatic stress disorder (PTSD), and artificial intelligence, announces the publication in the prestigious journal Scientific Reports (published by Nature Portfolio) of its intermediate results of a clinical study on major depressive disorder (MDD) demonstrating the existence of a predictive biosignature in the evolution of the disease with an average of 62 patient-specific physiological characteristics. The article, published on April 25th under the title "Predictive biosignature of major depressive disorder derived from physiological measurements of outpatients using machine learning" in open access on the Scientific Reports website, is the first scientific publication addressing this discovery in a peer-reviewed journal.
By 2030, MDD becomes the leading cause of disability and health burden worldwide. Globally, more than 280 million of people suffer from MDD1. When something difficult has happened, sad feelings and bad moods are a normal part of life. However, MDD goes beyond a feeling down: it affects the patient physically, emotionally, and cognitively. It is characterized by the following symptoms for at least 15 days2: a loss of interest and pleasure, a depressed mood, psychomotor agitation or retardation and fatigue or loss of energy nearly every day. The depressed patient may also experience anxiety, loss of appetite, insomnia or hypersomnia, low self-esteem and recurrent thoughts of death.
Psychiatrists and family doctors providing care for patients suffering from MDD have experienced the disease's heterogeneous characteristics, making difficult the prediction of its evolution and thus patient management.
"This innovation, supported by the US Office of Naval Research (ONRG), is an important result that will help bring the patient closer to his physician, leading to a more personalized patient management" comments Denis Fompeyrine, Ph.D. in Clinical Psychology, and Founder and CEO of MyndBlue. MDD is a highly unpredictable disease, and expressions of the disease's evolution is patient specific. This individual biosignature will become essential to anticipate the onset of crises in patients and ultimately increase the recovery rate.
The lack of personalized medical care leads to prescribing the same treatments for different types of depression: postpartum depression, insomnia depression, hypersomnia depression, and seasonal depression. However, MDD requires precision medicine a care that is adapted to each patient and the evolution of their disorders. In order to address this issue, MyndBlue has sought to develop a machine learning algorithm that identifies a biosignature to provide a clinical score of the depressive symptoms using individual physiological measurements.
"This study paves the way to digital phenotyping, using physiological multimodal approaches that I believe are necessary in the mood disorder clinical practice, explains Professor Pierre Alexis Geoffroy, a psychiatrist at the Bichat-Claude Bernard Hospital, AP-HP, Chronos Center, GHU Paris Psychiatry Neurosciences. And it is also thanks to the progress of artificial intelligence that it is possible to overcome all these limits. Our hope is for a precision medicine, as has been successfully developed for other diseases”.
An algorithm with the ability to predict the clinical status of the patient by a predictive biosignature
MyndBlue initiated a prospective, multicenter clinical trial in which outpatients diagnosed with major depressive disorder were enrolled and continuously wore a passive monitoring device for 6 months. Altogether, 150 physiological measures related to physical activity, heart rate, heart rate variability, respiratory rate, and sleep were captured.
For each patient, the algorithm was trained on daily physiological characteristics recorded during the first three months, as well as on corresponding standardized clinical assessments performed at baseline and during months 1, 2, and 3 by physicians. The algorithm's ability to predict the patient's clinical status was tested over the remaining 3 months.
The clinical trial's intermediate results reveal that the algorithm is able to characterize the physiological measures impacted by the patient's specific symptoms for each individual suffering from MDD and to assess the severity of the disease using these measures over a 3-month period with a sensitivity of 79%, and a specificity of 94%.
“These results allow the interpretation of all the physiological measurements as a biological signature of the disease, or biosignature, which characterizes the specificities of the phenotypic expression of the MDD” explains Nicolas Ricka, Ph.D. in mathematics and Head of AI Research at MyndBlue. “This discovery opens new research opportunities in the study and segmentation of patients suffering from MDD by objective and quantifiable criteria”.
This new categorization will allow physicians to conduct objective and personalized clinical follow-ups of their patients during and after consultations and intervene earlier thanks to remote real-time detection of complications and to predict the trajectory of the disease in order to anticipate relapse.
For more information, here is the link to access the publication:
MyndBlue is a digital MedTech company specializing in major depressive disorder, post-traumatic stress disorder, and artificial intelligence. MyndBlue is developing a personalized and predictive clinical score combining physiological measurements from outpatients, machine learning, and artificial intelligence.
MyndBlue's mission is to enable physicians to conduct objective clinical monitoring of their patients during and distance from consultations, intervene earlier through real-time detection of complications, and predict the trajectory of the disease to anticipate crisis and relapse. MyndBlue aims to provide patients with optimal treatment to increase their recovery rate and chances of regaining a high-quality life.
This press release contains, by implication or expressly, information and statements that may be considered prospective concerning MyndBlue.
The results presented in this publication are preliminary and may be re-evaluated at the end of the study, which is still on going. The medical device used in the study has not yet been approved for commercialization.
1 World Health Organization Fact Sheets—Depression. https://www.who.int/news-room/fact-sheets/detail/depression (2023)
2 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-5, 5th ed (American Psychiatric Publishing, a division of American Psychiatric Association, 2013).