Cardiogenic brain abscess
An analytical framework to identify predictive factors, and a protocol for early identification of brain abscess in patients with Tetralogy of Fallot (TOF).
Research
Research
Research at Mantrana starts with real questions from the field and the clinic. Social welfare analytics models how public health and welfare programmes work on the ground. In the hospital, some projects are already in development; others are being shaped with clinicians at SSSIHMS, where AI can make a difference in day-to-day hospital work.
In development
Applications in development
PTMC outcomes in mitral stenosis
A framework to predict outcomes and analyse risk factors for patients with mitral stenosis undergoing Percutaneous Transvenous Mitral Commissurotomy (PTMC).
Recurrent lumbar disc herniation
An explainable machine learning framework for predicting recurrent lumbar disc herniation.
Prof R Raghunath Sarma, Mr. M Subramanian, Dr. Sumit Thakar, Dr. Prayag Kini
Hospital AI projects
Projects in discussion with SSSIHMS
Ideas raised with SSSIHMS for applying AI, language models and automation to everyday hospital work, grouped by theme.
Radiology
4 projects
-
Triage radiology reports by flagging studies with urgent or emergency findings, and message the clinician and radiologist.
-
Identify image series of different body parts and file them in the correct folders in PACS.
-
Identify and monitor radiology key performance indicators (KPIs).
-
Voice-activated RIS for patient notes, placing orders, appointments, registration and statistics.
Scheduling and rosters
3 projects
-
Use language models to auto-schedule CT and MRI studies around the time each study needs and other constraints.
-
Use language models to build staff rosters that respect every constraint and leave detail.
-
Build duty rosters for radiology technicians from constraints described to the software.
Clinical documentation
1 project
-
Turn the doctor-patient conversation in the outpatient department into text and summarise it into OPD notes.
Patient communication
3 projects
-
Sort patient emails by department, with automatic replies for nephrology queries.
-
Send SMS alerts to ordering clinicians when lab reports are flagged with critical findings.
-
Let patients call in, enter their patient ID and hear whether their CT or MRI report is ready, through the hospital phone system.
Patient flow and operations
4 projects
-
Reduce patient waiting times for tests and for CT and MRI appointments.
-
Plan space and rooms for tests so that no area is crowded.
-
Design queuing for the urology outpatient department around its space, consulting rooms and patient flow.
-
Improve the efficiency of investigations together with the clinical departments.
Ask the data
1 project
-
Ask for statistics in plain language; the system understands the question and writes the database query to answer it.
Work with us
Students and researchers can contribute to these projects through internships.
Social welfare analytics
Social Welfare Analytics
Social welfare analytics applies systems modelling and AI to public health and welfare programmes. Each project brings together programme documents, survey evidence and feedback from the people who run and use a programme, to show where it reaches people, where it falls short, and which interventions would help.
See the 6 projects