Automated Detection and Classification of Schistocytes by a Novel Red Blood Cell Module Using Digital Imaging/Microscopy
Bibliographic record
Abstract
To the Editor The detection of morphological red blood cell (RBC) abnormalities is of vital importance in a variety of diseases. For thrombotic thrombocytopenic purpura (TTP), the presence of schistocytes in a peripheral blood smear is a major criterion for the diagnosis. TTP is an acute life-threatening disease in which the protein ADAM-TS13 (a disintegrin and metalloprotease with thrombospondin type 1 motif, member 13) is ineffective (e.g. due to the presence of antibodies to ADAM-TS13). The function of this protein is cleavage of the coagulation protein von Willebrand factor (VWF) into smaller pieces [1]. Since ADAM-TS13 is inactive, blood clots are formed more rapidly which capture thrombocytes, resulting in a thrombocytopenia. Erythrocytes are fragmented by these clots, leading to hemolytic anemia and schistocytes. Patients often present with fever, hemorrhage, neurological symptoms, petechia and renal damage. Alongside, an increased lactate dehydrogenase (LDH) and a decreased haptoglobin are seen. Schistocytes are morphologically polymorphic, which complicates the detection and identification. Also, a consistent and standardized international grading system for aberrant RBCs is still lacking nowadays [2]. The grading of abnormal RBCs is shown by a number of plus signs, which stands for the percentage of abnormal cells. In our laboratory, schistocytes are indicated with 1+, 2+ and 3+ (which stand for 0.5-2%, 2-5% and > 5% schistocytes) [3], but these percentages vary per laboratory. Recently a novel application module (Advanced RBC application, Cellavision) for the digital microscope was developed to automatically detect and classify morphological abnormalities of RBCs. This application actually works as a RBC-locating device. RBCs are identified and segmented by the system, after which each RBC is characterized for shape, size, color and inclusion. An artificial neural network is responsible for these characterizations, which is trained by a number of highly qualified experts. This neural network uses 80 features calculated for each RBC image. Some examples of used features are roundness, size, distribution of notches around the border and size and shape of inner pallor. The suggested morphology by the system can be changed or verified by the user. The results are represented as a percentage value and a grading which is based on a conversion table defined by the user in the settings file. With the use of digital imaging and this novel RBC mod
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".