From ff796954017f79769c813d6b4d619dafc57bd663 Mon Sep 17 00:00:00 2001
From: Jakub Kaczmarzyk <jakub.kaczmarzyk@gmail.com>
Date: Tue, 9 Jul 2024 14:15:48 -0400
Subject: [PATCH] [DOC] add disclaimer that this project is for academic
 research purposes only (#224) [skip ci]

* add academic research use warning to readme

* add academic research disclaimer

* update float and int types for slide coords and probs

Use np.floating and np.integer.

* do not ignore types when making binary im arr
---
 README.md                         | 3 +++
 docs/index.rst                    | 4 ++++
 wsinfer/modellib/run_inference.py | 4 ++--
 wsinfer/patchlib/segment.py       | 5 +----
 4 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/README.md b/README.md
index c8e4f1e..9d71e49 100644
--- a/README.md
+++ b/README.md
@@ -12,6 +12,9 @@ Original H&E                        |  Heatmap of Tumor Probability
 [![Supported Python versions](https://img.shields.io/pypi/pyversions/wsinfer)](https://pypi.org/project/wsinfer/)
 [![Published in npj Precision Oncology](https://img.shields.io/badge/Published-npj_Precision_Oncology-blue)](https://doi.org/10.1038/s41698-024-00499-9)
 
+> [!CAUTION]
+> WSInfer is an academic project intended for research use only.
+
 See https://wsinfer.readthedocs.io for documentation.
 
 The main feature of WSInfer is a minimal command-line interface for running deep learning inference
diff --git a/docs/index.rst b/docs/index.rst
index fa9280f..6f0b054 100644
--- a/docs/index.rst
+++ b/docs/index.rst
@@ -23,6 +23,10 @@ WSInfer: blazingly fast inference on whole slide images
 on whole slide images. It includes several built-in models, and it can be used with any
 PyTorch model as well. The built-in models :ref:`are listed below <available-models>`.
 
+.. caution::
+
+  WSInfer is an academic project intended for research use only.
+
 Running inference on whole slide images is done with a single command line:
 
 ::
diff --git a/wsinfer/modellib/run_inference.py b/wsinfer/modellib/run_inference.py
index 7ba8aa5..0a0926e 100644
--- a/wsinfer/modellib/run_inference.py
+++ b/wsinfer/modellib/run_inference.py
@@ -171,8 +171,8 @@ def run_inference(
 
         # Store the coordinates and model probabiltiies of each patch in this slide.
         # This lets us know where the probabiltiies map to in the slide.
-        slide_coords: list[npt.NDArray[np.int_]] = []
-        slide_probs: list[npt.NDArray[np.float_]] = []
+        slide_coords: list[npt.NDArray[np.integer]] = []
+        slide_probs: list[npt.NDArray[np.floating]] = []
         for batch_imgs, batch_coords in tqdm.tqdm(loader):
             assert batch_imgs.shape[0] == batch_coords.shape[0], "length mismatch"
             with torch.no_grad():
diff --git a/wsinfer/patchlib/segment.py b/wsinfer/patchlib/segment.py
index a6963fd..cd071e4 100644
--- a/wsinfer/patchlib/segment.py
+++ b/wsinfer/patchlib/segment.py
@@ -66,10 +66,7 @@ def segment_tissue(
     _, im_arr = cv.threshold(
         im_arr, thresh=binary_threshold, maxval=255, type=cv.THRESH_BINARY
     )
-
-    # Convert to boolean dtype. This helps with static type analysis because at this
-    # point, im_arr is a uint8 array.
-    im_arr_binary: npt.NDArray[np.bool_] = im_arr > 0  # type: ignore
+    im_arr_binary: npt.NDArray[np.bool_] = im_arr > 0
 
     # Closing. This removes small holes. It might not be entirely necessary because
     # we have hole removal below.