This Lucknow-based startup is using AI to deliver quality diagnostics across India

For millions of Indians, getting an accurate medical diagnosis remains out of reach. When a patient experiences persistent breathing problems or chest pain, they need specialised tests and expert analysis to understand what's wrong. 

Yet India's diagnostic system is "highly fragmented and largely unregulated," according to a 2015 PLOS ONE study, with most testing carried out in small, ill-equipped laboratories scattered across the public and private sectors.

Siblings Ankit and Saumya Shukla, co-founders of diagnostics solutions company Dectrocel, saw this problem up close. Their parents worked across medical colleges, and they witnessed how patients struggled to get accurate diagnoses.

Ankit pursued a PhD in AI and Machine Learning in Applied Medicine from the University of Queensland, collaborating with Google Brain on drug-resistant tuberculosis. Saumya earned a Doctorate in Public Health from Sanjay Gandhi Post Graduate Institute of Medical Sciences (SGPGIMS) and studied data science at Harvard University. This combined expertise allowed them to fill the critical gap they identified.

According to Ankit, India has only one radiologist for every 125,000 people. "In Tier II cities and smaller towns, diagnosis delays stretch to 7-10 days with error rates between 25-40%," he says. "Even when remote clinics have chest X-ray machines, there's often no radiologist to read them."

The result? Patients get their X-rays taken, but struggle to get accurate interpretations. Those with complex respiratory conditions often wait months, sometimes over a year, before receiving an accurate diagnosis. "By the time some patients are correctly diagnosed, treatment options become very limited," says Saumya.

Diagnostic variability only complicates it further. "Doctors don't always agree on what they see in scans," Ankit notes, leading patients to seek multiple opinions and delaying treatment further.

The DecXpert platform

This became the foundation for Dectrocel. Founded in 2020 in Lucknow by Ankit Shukla, Saumya Shukla, and Nikhil Mishra, Dectrocel develops AI-driven diagnostic solutions for radiology. Mishra leads the company as CTO and is an IIT Kanpur researcher specialising in AI and machine learning in medicine. The trio built a solution that uses AI to analyse chest X-rays accurately, bringing specialised diagnostic capabilities to healthcare facilities across India. 

The flagship product, DecXpert, is a clinical decision support platform that reads and interprets chest X-rays and CT scans. The chest X-ray module analyses scans for critical respiratory conditions, including lung cancer, pneumonia, tuberculosis, and pneumothorax. 

The CT scan module provides deeper analysis for complex cases. When an X-ray suggests possible lung cancer, the CT scan AI can determine the cancer's stage, evaluate conditions like Interstitial Lung Disease, and provide detailed measurements of lung lesions.

DecXpert generates standardised reports that identify specific conditions, pinpoint affected lung areas, and include visual guidance highlighting areas of concern. The system includes a review protocol where cases can be escalated to senior doctors when consulting physicians disagree with the AI's findings.

"Our tuberculosis diagnosis is benchmarked against molecular PCR tests like GeneXpert, the gold standard for TB detection," Saumya explains. "While PCR tests can take days, our AI provides immediate results with 98% accuracy, 10x faster than traditional methods."

DecXpert is one of only three AI-based diagnostic tools approved by the Central Drugs Standard Control Organisation (CDSCO), India's medical device regulator, and is ISO 13485-certified for quality management.

How it works

Dectrocel's software integrates directly with existing X-ray, CT, MRI, and PET-CT machines, aiming to eliminate the need for additional hardware. For equipment manufacturers (OEMs), the integration can happen at the machine level, while clinics can integrate it directly into their existing systems.

When a scan is taken, the system generates two instant outputs: the standard scan image and an AI-generated diagnostic report as a digital PDF, which can be sent to patients via WhatsApp or other channels.

The platform works within hospital PACS (Picture Archiving and Communication System), fitting into existing clinical workflows. Patients can also upload chest X-rays directly through Dectrocel's WhatsApp, website, or app for analysis.

Building trust through data

Behind that 98% accuracy lies nearly a decade of R&D. Dectrocel's AI was trained on chest X-rays from 1 million patients and tested on another 500,000 patients across 24 hospitals in North and South India, including SGPGI Lucknow,  Google Brain, and Apollo Hospitals. 

Government health centre pilots in Gorakhpur and Kanpur, along with BP Singh Hospital, Midland Lucknow, and Vivekananda Hospital, showed a 40% increase in respiratory disease detection. “This means thousands of previously missed cases were finally identified,” says Saumya. "For public health screening, this is a massive leap," she adds.

The validation, conducted alongside 50 senior specialists from SGPGI's Department of Pulmonary Medicine, was peer-reviewed and published in Nature Scientific Reports with additional review by WHO UK evaluators. To date, the technology has screened over 90,000 patients across India, many through rural health camps.

Dectrocel competes with established players like Qure.ai and DeepTek. It aims to differentiate itself through its system design: it runs on local servers rather than cloud services, eliminating foreign exchange costs and data privacy concerns. The system works with any X-ray equipment, from local Indian manufacturers to international brands.

"This makes our pricing four to five times lower than competitors," Ankit says. 

Funding and growth plans

Dectrocel raised Rs 4 crore in its seed funding round led by IAN Group's BioAngels, with participation from PadUp Ventures and Vinners, with angel investors including Nitin Zamre, Samir Kalia, and Mitesh Shah.

The funding will fuel commercial rollout, OEM integrations, and expansion into Southeast Asia. Product development will focus on new diagnostic modules for CT, MRI, and PET-CT scans, as well as modules for hepato-pancreato-biliary (HPB) diseases. 

The company is also working with the All India Institute of Medical Sciences and Maulana Azad Medical College to develop AI algorithms for oncology applications in PET-CT and MRI imaging.

The founding team's core mission remains the same despite expansion plans. "We're not just building AI diagnostic tools, we're saving lives by ensuring that a patient in a Tier III city gets the same quality interpretation as someone in a metro hospital," Ankit says. "Our 40.1% improvement in missed case detection, validated in Nature Scientific Reports, demonstrates that AI can be the great equaliser in healthcare access."