AI SQUASH
NIMHANS Bangalore
Imagine a world where the tension of waiting for crucial brain tumor pathology reports during surgery is a thing of the past. That’s exactly what AI SQUASH, a pioneering startup founded by a group of doctors from NIMHANS, is working towards. Their journey began with a significant challenge: during neurosurgical procedures, the need for rapid, accurate tumor analysis often leads to delays of up to an hour. These delays not only waste valuable operating room time but also prolong patient anesthesia and can impact surgical outcomes.
AI SQUASH is addressing this problem with a sophisticated AI-driven solution. Over the past three years before joining I-NCUBATE Program, the team developed an advanced convolutional neural network (CNN) model, trained on thousands of tumor images, to provide rapid and precise predictions of tumor grades. This technology offers results within minutes, achieving accuracy comparable to experienced neuropathologists.
Initially, the team targeted all neurosurgeons, believing their solution would be universally beneficial. However, through GDC’s I-NCUBATE Program, it became clear that their approach needed refinement. They discovered that not all neurosurgeons encounter significant delays that would warrant their AI solution. For many, especially those with reliable frozen or squash biopsy options, the impact of delayed reporting was minimal.
A pivotal moment in their journey occurred when they engaged with a neurosurgeon, who highlighted the severe delays experienced during stereotactic biopsy procedures. This feedback illuminated a crucial insight: while some neurosurgeons faced minimal consequences from reporting delays, others, particularly those handling STB biopsies, encountered significant disruptions. This realization led AI SQUASH to pivot their focus, homing in on these high-impact scenarios where their technology could make the most difference.
With GDC’s expert mentorship, AI SQUASH has refined its strategy to address these specific needs. The startup is now concentrating on improving efficiency for STB biopsies and exploring additional applications, such as tumor edge clearance and slide screening for pathologists.
