One row per modality: the hollow dot is 2024 papers, the solid dot is 2025, and the connector between them is the change. Rows are ordered by the size of that change in papers, not by percentage.
PubMed papers per modality, 2024 and 2025
papers published in the year
Of the 19 modalities with at least 20 papers, 17 higher, 1 lower and 1 unchanged; the largest single move is Botulinum toxin (cosmetic), up 52 papers.
One hue at two shades, because both ends are the same quantity for the same modality a year apart. Direction is never colour alone: every row prints its signed change with an arrow. 6 rows drawn at reduced opacity carry fewer than 20 papers, so the spread shown is computed from very few records. The exact count is printed on every row.
PubMed papers per modality in 2024 and 2025, and the change between them
2024
2025
Change
Botulinum toxin (cosmetic)
206
258
+52
Poly-L-lactic acid collagen stimulator
65
109
+44
Ablative / hybrid fractional laser
94
134
+40
Hyaluronic acid filler
164
196
+32
Calcium hydroxylapatite filler
64
94
+30
Non-ablative fractional laser
82
111
+29
Mechanical microneedling
92
119
+27
Ultrasound lifting & tightening
37
64
+27
Monopolar RF skin tightening
37
58
+21
IPL / broadband light
49
65
+16
Picosecond laser (pigment & tattoo)
38
51
+13
Pulsed dye laser
78
88
+10
RF body contouring
26
36
+10
Cryolipolysis
12
21
+9
Topical retinoid
162
169
+7
Radiofrequency microneedling
44
49
+5
Topical clascoterone
18
21
+3
Sebaceous-gland laser for acne
7
8
+1
Bimatoprost for eyelash growth
7
7
0
Topical hydroquinone / triple combination
70
70
0
Non-invasive laser lipolysis
4
3
−1
High-intensity electromagnetic muscle stimulation
8
5
−3
Laser hair removal
50
46
−4
Hydradermabrasion
9
2
−7
Cellulite septa release
10
2
−8
Two fetched year counts per modality, drawn as they are — no index, no baseline, no fitted trend. What this measuresEach row plots that modality's PubMed count for 2024 and for 2025, the two most recent FULL years; the partial current year is not on this chart at all. A modality missing either year is drawn as a worded row rather than joined to zero. Rows whose larger year is under 20 papers are drawn at reduced opacity, because a move of three papers on a base of six is not the same event as a move of three papers on a base of six hundred, and this chart cannot show the difference in its geometry. Number of PubMed records matching each modality's fixed, hand-written title/abstract query, windowed to a single publication year. It measures publication volume about a TECHNIQUE, not a brand. It is NOT a quality score, NOT an efficacy measure, and NOT a count of randomised trials. 2026 is a partial year (run date 2026-07-26); its bar is short because the year is short, not because interest fell. 2026 contains only papers indexed by the run date (2026-07-26). Render it visually distinct (e.g. hollow/hatched) or annotate it; never let it read as a decline. Methodology →
Year-over-year change in PubMed publication volume for all 25 tracked modalities, 2025 vs 2024, with both base counts.
ModalityModalityA family of treatments that work the same way — such as radiofrequency microneedling, or lasers that leave the skin surface intact — rather than any one branded machine.
Category
2024 papersPubMedThe United States National Library of Medicine's index of published medical literature, which is where this site's paper counts come from.
2025 papers
Change in papers
Change, per cent
2016–2026Partial yearA year that has not finished, so its count covers only the part of the year already on record.
Ordered by ABSOLUTE change, because a percentage on a base of four is noise, not momentum. What this measuresThe table ranks by the change in the number of papers, not by percentage, for one reason: a modality that went from 4 papers to 8 is +100% and a modality that went from 400 to 480 is +20%, and only one of those is a real shift in research attention. Both base counts are printed in their own columns so you can always see what a percentage was computed from, and the percentage chip renders an em-dash rather than a number when the 2024 base is zero or missing — a division by zero is not a trend. Number of PubMed records matching each modality's fixed, hand-written title/abstract query, windowed to a single publication year. It measures publication volume about a TECHNIQUE, not a brand. It is NOT a quality score, NOT an efficacy measure, and NOT a count of randomised trials. 2026 is a partial year (run date 2026-07-26); its bar is short because the year is short, not because interest fell. 2026 contains only papers indexed by the run date (2026-07-26). Render it visually distinct (e.g. hollow/hatched) or annotate it; never let it read as a decline. Methodology →
Under every card: papers published in 2025, the last full year, then the change against 2024. Every card shares one vertical scale, so a flat-looking line means a small series, not a stable one.
Sebaceous-gland laser for acne — PubMed papers per year, 2016–2026 (2026 partial)
Sebaceous-gland laser for acne
2016
2
2017
3
2018
0
2019
4
2020
2
2021
1
2022
5
2023
7
2024
7
2025
8
2026
10
8↑+14%
Every card shares ONE y-axis, so a flat-looking line means a small series, not a stable one. What this measuresEach card counts PubMed records per year for that modality query. It is a count of publications, not a measure of quality, effect size or clinical outcome, and a paper is counted once whether it is a randomised trial or a single case report. All 18 cards are drawn on a single shared y-axis maximum (258 papers, the largest annual value across every modality on this page) so heights can be compared card to card; the cost of that choice is that a genuinely small series renders as a nearly flat line rather than filling its own box. 2025 is the last full year and 2024 is its comparison base. The final point of every sparkline is the partial current year and is drawn hollow for that reason. The percentage chip renders an em-dash rather than a number when the 2024 base is zero or missing, because a division by zero is not a trend. Number of PubMed records matching each modality's fixed, hand-written title/abstract query, windowed to a single publication year. It measures publication volume about a TECHNIQUE, not a brand. It is NOT a quality score, NOT an efficacy measure, and NOT a count of randomised trials. 2026 is a partial year (run date 2026-07-26); its bar is short because the year is short, not because interest fell. 2026 contains only papers indexed by the run date (2026-07-26). Render it visually distinct (e.g. hollow/hatched) or annotate it; never let it read as a decline. Methodology →
Under every card: papers published in 2025, the last full year, then the change against 2024. Every card shares one vertical scale, so a flat-looking line means a small series, not a stable one.
Every card shares ONE y-axis, so a flat-looking line means a small series, not a stable one. What this measuresEach card counts PubMed records per year for that modality query. It is a count of publications, not a measure of quality, effect size or clinical outcome, and a paper is counted once whether it is a randomised trial or a single case report. All 5 cards are drawn on a single shared y-axis maximum (258 papers, the largest annual value across every modality on this page) so heights can be compared card to card; the cost of that choice is that a genuinely small series renders as a nearly flat line rather than filling its own box. 2025 is the last full year and 2024 is its comparison base. The final point of every sparkline is the partial current year and is drawn hollow for that reason. The percentage chip renders an em-dash rather than a number when the 2024 base is zero or missing, because a division by zero is not a trend. Number of PubMed records matching each modality's fixed, hand-written title/abstract query, windowed to a single publication year. It measures publication volume about a TECHNIQUE, not a brand. It is NOT a quality score, NOT an efficacy measure, and NOT a count of randomised trials. 2026 is a partial year (run date 2026-07-26); its bar is short because the year is short, not because interest fell. 2026 contains only papers indexed by the run date (2026-07-26). Render it visually distinct (e.g. hollow/hatched) or annotate it; never let it read as a decline. Methodology →
Neither a riser nor a faller
2 modalities
Flat and unknown are kept visible, not dropped — a missing series is not a zero. What this measures2 modality series recorded exactly the same count in both years (Bimatoprost for eyelash growth, Topical hydroquinone / triple combination). These are excluded from the riser and faller grids because they have no direction to show, not because they were filtered for looking uninteresting. Every one of them is still in the table above with its real numbers. Methodology →