Tissue staining quality determination

An image analysis method for automatically determining, for one or more staining protocol parameters of an IHC staining protocol, a respective parameter value range that will result in a staining of a tissue sample with a tumor-marker specific stain that fulfills predefined staining quality requirements, the method comprising:

• receiving a plurality of images of one or more IHC stained tissue samples belonging to multiple different extended tissue types, the pixel intensities of each image correlating with the amount of a tumor-marker-specific stain, the depicted tissue samples having been stained with the same tumor-marker-specific stain but with different values of each of the one or more staining protocol parameters;
• extracting a plurality of features from each of the received digital images, wherein the extracted plurality of features comprises at least information relating to the pixel intensities of the digital image;
• automatically analyzing the extracted features for identifying an extended tissue type using a first classifier and a contrast level of each of the tissue samples using a second classifier, wherein the extended tissue type comprises a defined expression level of a presence of the tumor marker and the identification of the extended tissue type includes an identification of the defined expression level by at least:
• inputting the extracted features of the respective received image into the first classifier, the first classifier being configured to identify the extended tissue type of the tissue depicted in the digital image as a function of at least some first ones of the extracted features of the respective received image;
• and
• inputting the extracted features of the respective received image into the second classifier, the second classifier being configured to identify the contrast level of the tissue depicted in the digital image as a function of at least some second ones of the extracted features of the respective received image, the contrast level indicating the intensity contrast of pixels of the stained tissue;
• for the tissue depicted in each of the received images, computing a staining quality score as a function of the identified extended tissue type from the first classifier and the identified contrast level of the tissue from the second classifier;
• performing, for each of the one or more staining protocol parameters:
• for each of the extended tissue types:
• identifying all images depicting a tissue of said extended tissue type for which a quality score that exceeds or meets an acceptable quality threshold was computed;
• identifying the minimum and maximum staining parameter value used in the staining protocols that were applied for staining the tissues depicted in said identified images;
• returning a staining parameter value range for the staining protocol parameter, the range being delimited by the identified minimum and maximum values.

An image analysis system comprising:

• a storage medium comprising a first classifier and a second classifier;
• an interface configured for receiving a digital image of an IHC stained tissue, the pixel intensities of the image correlating with the amount of a tumor-marker-specific stain;
• a processor configured for:
• extracting a plurality of features from the received digital image, wherein the extracted plurality of features comprises at least information relating to the pixel intensities of the digital image;
• inputting the extracted features into the first classifier, the first classifier being configured to identify an extended tissue type of the tissue depicted in the digital image as a function of at least some first ones of the extracted features, wherein the extended tissue type comprises a defined expression level of a presence of the tumor marker and the identification of the extended tissue type includes an identification of the defined expression level;
• inputting the extracted features into the second classifier, the second classifier being configured to identify a contrast level of the tissue depicted in the digital image as a function of at least some second ones of the extracted features, the contrast level label indicating the intensity contrast of pixels of the stained tissue;
• computing a staining quality score for the tissue depicted in the digital image as a function of the identified extended tissue type from the first classifier and the identified contrast level from the second classifier.

An image analysis system comprising:

• a storage medium comprising a first classifier and a second classifier;
• an interface configured for receiving a plurality of images of one or more IHC stained tissue samples belonging to multiple different extended tissue types, the pixel intensities of each image correlating with the amount of a tumor-marker-specific stain, the depicted tissue samples having been stained with the same tumor-marker-specific stain but with different values of each of the one or more staining protocol parameters;
• a processor configured for:
• extracting a plurality of features from each of the received digital images, wherein the extracted plurality of features comprises at least information relating to the pixel intensities of the digital image;
• automatically analyzing the extracted features for identifying an extended tissue type using the first classifier and a contrast level of each of the tissue samples using the second classifier, wherein the extended tissue type comprises a defined expression level of a presence of the tumor marker and the identification of the extended tissue type includes an identification of the defined expression level by at least:
• inputting the extracted features of the respective received image into the first classifier, the first classifier being configured to identify the extended tissue type of the tissue depicted in the digital image as a function of at least some first ones of the extracted features of the respective received image;
• and
• inputting the extracted features of the respective received image into the second classifier, the second classifier being configured to identify the contrast level of the tissue depicted in the digital image as a function of at least some second ones of the extracted features of the respective received image, the contrast level indicating the intensity contrast of pixels of the stained tissue;
• for the tissue depicted in each of the received images, computing a staining quality score as a function of the identified extended tissue type from the first classifier and the identified contrast level of the tissue from the second classifier;
• performing, for each of the one or more staining protocol parameters:
• for each of the extended tissue types:
• identifying all images depicting a tissue of said extended tissue type for which a quality score that exceeds or meets an acceptable quality threshold was computed;
• identifying the minimum and maximum staining parameter value used in the staining protocols that were applied for staining the tissues depicted in said identified images;
• returning a staining parameter value range for the staining protocol parameter, the range being delimited by the identified minimum and maximum values, the range being a staining parameter range adapted to stain a biological sample with a tumor-marker specific stain such that predefined staining quality requirements are fulfilled.