Problem Statement

The project addressed the issue of inter-reader variability among radiologists in evaluating lumbar spine MRI images for spinal stenosis and disc degeneration. By developing a PDF report prototype with a fully automated software solution, the goal was to generate a quantitative approach that simplifies the workflow for clinicians, reduces their workload, and enhanced the services provided by medical practices.

Definition

The project involved developing a PDF Report for evaluating lumbar spine MRI images using AI-based software to quantify spinal stenosis and disc degeneration. The aim was to create a user-friendly and efficient tool that decreases inconsistency among radiologists’ assessments, thus improving throughput in medical routines and attracting potential clients to the differentiated services.

Data Analysis

The project required comprehensive analysis of a diverse dataset of lumbar spine MRI images. The data architecture involved organizing, manually marking and training machine learning algorithms and preprocessing the images to ensure accurate evaluations. The software procesed the data to generate objective and reliable assessments, contributing ot reduced variability in radiologists’ readings.