What is actually in this field?

A technologist scanning a urine sediment at 400x sees a cluster of small round structures near a group of red blood cells (RBCs). Two of them look like they might have a faint bump. A reagent-strip result on the same specimen shows blood 2+ and trace leukocyte esterase (LE). Nothing about the field is obviously wrong, but nothing is obviously settled either, and the report has to say something specific: RBC, white blood cell (WBC), epithelial cell, yeast, bacteria, or artifact.

This is the ordinary bench problem behind image-atlas work. Urine sediment holds a mix of true cellular elements, organisms, and non-biological debris that overlap in size and can share a rounded outline under bright-field optics. A single static image rarely settles the question. What separates a confident call from a guess is a small, repeatable set of features: relative size against a known reference, border quality, internal detail, behavior across focus planes, and correlation with the chemical strip.

The guided case is a clean-catch specimen received 35 minutes after collection, with specific gravity 1.008, pH 8.0, blood 2+, LE trace, nitrite negative, 1-3 RBCs per high-power field (HPF) with many pale rings, 4-7 WBCs/HPF, many squamous epithelial cells, and 8-12 small round refractile structures per HPF.

The same field sorts into confident identification, probable category, or requires review, with two visible features named for each call. The mental model and feature checklist make that sorting defensible rather than guessed.