Nile Wilson

For some pre-defined recognizers, a default value (e.g., 0.85) is assigned for the NER score when the score is not provided. For example in https://github.com/microsoft/presidio/blob/main/presidio-analyzer/presidio_analyzer/predefined_recognizers/spacy_recognizer.py Having a valid number hard-coded...

**Describe the bug** The `DicomImageRedactorEngine` class currently only supports greyscale (i.e., MONOCHROME1 and MONOCHROME2) and RGB [photometric interpretations](https://www.dicomobjects.com/kb/Photometric-Interpretations/#:~:text=Photometric%20Interpretations%20DICOM%20allows%20various%20relationships%20between%20the,0028%2C0004%20and%20every%20image%20should%20have%20that%20value.). The class should be updated to ensure DICOM images of different [photometric...

**Is your feature request related to a problem? Please describe.** Multiple points of code in the `presidio-image-redactor` module rely on the OCR results fitting a certain format. While this works...

**Describe the bug** When running redaction on compressed pixel data, the returned pixel data is uncompressed. This is because when adding boxes via `DicomImageRedactorEngine._add_redact_box`, we use the loaded DICOM instance's...

**Describe the bug** When calling `.verify_dicom_instance` using the `DicomImagePiiVerifyEngine`, the returned verification image is empty when no PII text is detected. **To Reproduce** Steps to reproduce the behavior: 1. Open...

**Is your feature request related to a problem? Please describe.** The `DicomImageRedactorEngine` and `DicomImagePiiVerifyEngine` can be instantiated and run locally, but there is currently no support for API calls to...

Inpaint instead of using solid color for bounding boxes

1

**Is your feature request related to a problem? Please describe.** Currently, every redactor box drawn is a solid color. The color is set to either contrast with or blend in...

Add ability to auto-detect which DICOM files to process

1

**Is your feature request related to a problem? Please describe.** When working with DICOM image datasets, the number of images that have text PHI burnt into them are usually fairly...

**Is your feature request related to a problem? Please describe.** Currently, image redaction and image PII verification operate under two separate classes in the `presidio-image-redactor` module. Having two separate imports...

**Is your feature request related to a problem? Please describe.** The `AnalyzerEngine` returns False Positive entities (e.g., labeling non-PII text as PII) with fairly high scores (e.g., above 0.7). **Describe...