Moscow City Medical Decision Support System

2021: Moscow therapists choose the diagnosis proposed by the neural network in 50% of cases

In July 2021, the Moscow authorities reported that therapists in metropolitan clinics choose the diagnosis proposed by the neural network in almost half of cases. We are talking about the use of a system for supporting medical decision-making, which was launched in all polyclinics in cities in September 2020.

From that moment to July 2021, the system became 7.4% more accurate due to the fact that it studied for 12 million visits. When choosing one or three diagnoses, the doctor agrees with the proposals of artificial intelligence in 46% and 68% of cases, respectively, said Anastasia Rakova, deputy mayor of Moscow for social development.

Therapists in Moscow choose the diagnosis proposed by the neural network in half of cases

According to her, the development of digital services in healthcare as a whole contributes to reducing the likelihood of error when making a preliminary diagnosis and increasing the detection of diseases at the primary level. The development of a system for supporting medical decision-making minimizes the routine in the actions of a doctor and allows him to pay attention to details that could elude the field of view. It also avoids being re-assigned to the same study.

The medical decision support system is represented by two modules. The pre-diagnosis module analyzes the patient's complaints and selects one of the three most possible diagnoses based on them. At the same time, the doctor can both choose one of the proposed diagnoses and put his own, as well as prescribe additional studies or referrals for consultations, as well as remove from the list those that the patient recently underwent.

The neural network processes the text of complaints submitted by the doctor, compares it with millions of depersonal patient records at EMIAS, analyzes patterns and offers the doctor the three most likely diagnoses.^[1]

Notes