<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Microscopy | Shovito Barua Soumma</title><link>https://www.shovitobarua.com/tag/microscopy/</link><atom:link href="https://www.shovitobarua.com/tag/microscopy/index.xml" rel="self" type="application/rss+xml"/><description>Microscopy</description><generator>Wowchemy (https://wowchemy.com)</generator><language>en-us</language><copyright>© 2025 Shovito Barua Soumma</copyright><lastBuildDate>Sat, 20 Jun 2026 00:00:00 -0500</lastBuildDate><image><url>https://www.shovitobarua.com/media/sharing.png</url><title>Microscopy</title><link>https://www.shovitobarua.com/tag/microscopy/</link></image><item><title>Bioimage QC Agent</title><link>https://www.shovitobarua.com/tool/bioimage-qc-agent/</link><pubDate>Sat, 20 Jun 2026 00:00:00 -0500</pubDate><guid>https://www.shovitobarua.com/tool/bioimage-qc-agent/</guid><description>&lt;img src="dashboard.gif" alt="GCSD dashboard demo" style="width: 100%; border-radius: 6px; margin-bottom: 1.5rem;">
&lt;p>Bioimage QC Agent is an interactive microscopy image segmentation and quality-control platform designed to improve the reliability and reproducibility of bioimage analysis workflows. The system integrates classical watershed segmentation, Cellpose deep-learning models, and the Segment Anything Model (SAM / micro-SAM) with an agentic quality-control framework that automatically detects segmentation failures and recommends corrective actions.&lt;/p>
&lt;p>The platform follows an observe → decide → act → evaluate loop. Segmentation outputs are analyzed using image-quality metrics, object statistics, and benchmarking measures to identify over-segmentation, under-segmentation, low contrast, blur, and other common failure modes. When issues are detected, the QC agent recommends parameter adjustments and can automatically rerun segmentation to improve results.&lt;/p>
&lt;p>A complementary Vision Transformer (ViT)-based quality-control classifier provides a learned alternative to rule-based QC assessment, enabling automated image-quality prediction without requiring manually annotated QC datasets.&lt;/p>
&lt;p>Bioimage QC Agent supports:&lt;/p>
&lt;ul>
&lt;li>Watershed-based nuclei segmentation&lt;/li>
&lt;li>Cellpose deep-learning segmentation&lt;/li>
&lt;li>Segment Anything Model (SAM) and micro-SAM segmentation&lt;/li>
&lt;li>Automated segmentation benchmarking (Dice, IoU, object counts)&lt;/li>
&lt;li>Rule-based quality-control assessment&lt;/li>
&lt;li>ViT-based QC classification&lt;/li>
&lt;li>Agentic parameter correction and rerun optimization&lt;/li>
&lt;li>Interactive visualization and reporting through a Streamlit dashboard&lt;/li>
&lt;/ul>
&lt;p>The platform was developed using the BBBC038 / 2018 Data Science Bowl nuclei segmentation benchmark dataset and demonstrates practical agentic AI for biomedical image analysis.&lt;/p>
&lt;p>The web dashboard allows users to:&lt;/p>
&lt;ul>
&lt;li>Upload fluorescence or brightfield microscopy images&lt;/li>
&lt;li>Interactively tune segmentation parameters&lt;/li>
&lt;li>Compare raw images, masks, overlays, and object statistics&lt;/li>
&lt;li>Inspect QC metrics and segmentation performance&lt;/li>
&lt;li>View agent recommendations and automatic corrections&lt;/li>
&lt;li>Download masks, overlays, metrics, and QC reports&lt;/li>
&lt;/ul>
&lt;p>
&lt;i class="fas fa-external-link-alt pr-1 fa-fw">&lt;/i> &lt;a href="https://bioimage-qc.streamlit.app/" target="_blank" rel="noopener">Open Tool&lt;/a>&lt;/p></description></item></channel></rss>